Abstract 19336: Reported Stroke Rates in Patients With Nonvalvular Atrial Fibrillation in North America are Lower Than Those Cited in ESC and AHA Guidelines

Abstract

Background: Anticoagulants (OAC) decrease ischemic stroke rates in patients with atrial fibrillation (AF) but increase the risk of bleeding. The risk of ischemic stroke where the absolute benefit of anticoagulation outweighs the bleeding risk has been shown to be 1-2% per year. The American Heart Association/American College of Cardiology/Heart Rhythm Society and the European Society of Cardiology (ESC) guidelines have adopted the CHA2DS2-VASc stroke risk score; their recommendations assume the scheme’s point scores correspond to fixed stroke rates. However, reported rates of stroke vary widely across cohorts, placing in question the generalizability of guideline recommendations. Objective: To contrast the reported rates of stroke in North American (NA) patients with AF who do not take OAC with the Danish AF cohort used to create the ESC guidelines. Methods: We conducted a systematic review to identify all cohort studies and randomized controlled trials including patients with non-valvular AF not treated with OAC. We excluded studies that enrolled only patients undergoing surgical procedures or cardioversion, or only patients with specific comorbidities such as prior stroke or kidney disease. Results: Of the 3,552 studies screened, we identified 13 eligible NA studies representing 137,652 patients. Larger and more contemporary cohorts generally had lower stroke rates (Table). When weighted by number of subjects, the NA cohorts’ ischemic stroke rate averaged 1.88% per year, whereas the Danish cohort had a rate of 4.66% per year. Conclusions: Large variation exists in stroke rates across putatively representative cohorts. This may reflect true differences in rates or methodologic differences among studies. These differences could change the point score threshold for recommending OAC in lower risk regions. Reexamining the net benefit of OAC in patients with a CHA2DS2-VASc score of 1 or 2 seems warranted, particularly for those with weak 1-point risk factors.